Genetic mutations can result from polymerase errors on both undamaged and chemically damaged DNA templates, potentially causing human disease including cancer. Despite substantial efforts from many laboratories, no clear picture has yet emerged to explain the mechanism of base misincorporation and mutagenesis. In this competitive renewal application, we propose that equilibrium properties of aberrant base pairs may hold the key to understanding base mispair formation and extension past the lesion as outlined in the progress report. The current proposal is divided into four specific aims. In aim 1, we propose to examine the role of the sugar pucker of the 3' terminal base residue of the DNA primer strand in extension past mispairs. Previous studies have shown that the inefficiency of mispair extension contributes as much to mutation avoidance as does discrimination at the base insertion step; however, the mechanism for such discrimination is as yet unknown. We have prepared a series of fluorine-substituted sugar analogs in which the conformation of the sugar of the 3'-terminal base is constrained. Preliminary studies with constrained analogs demonstrate that the sugar pucker is an important determinant in polymerase extension. We propose to study a series of Such analogs with several replicative and error prone DNA polymerases. In aim 2, we propose to study the structures of the 3'-terminal base pairs in template-primer complexes examined in aim 1 in solution using novel isotope-edited NMR methods in order to confirm the correlation between structural and biochemical perturbations. In aim 3, we propose to investigate relationships between characteristic mutation patterns established for various DNA polymerases and mutation spectra observed in the p53 tumor suppressor gene from human tumor samples. Several previous studies have established correlations between specific forms of DNA damage arldcharactedstic p53 mutations; however, the mutation spectra from many common human tumor types remain unexplained. We found unexpectedly that the p53 mutation spectra of several human tumor types can be explained by polymerase alpha and beta error frequencies. We propose to extend these studies to include additional polymerases. In aim 4, we propose to study 5-chloropyrimidines as potential markers of accumulated endogenous damage. The data obtained in the studies proposed here will contribute substantially to understanding the mechanisms responsible for mutations in human tumors and lead to rational strategies directed at minimizing genetic instability.